Which is Better?

[spyderman]

I'm in the process of designing a relational model for a database which will store petroleum data. I will need to track data for the following objects. Below, in uppercase, are some of the objects with their associated properties list below them:

WELL

- well name
- liquid rate
- gas rate
- shut in pressure
- flowing pressure
- CO2
- N2

NODE
- node name
- fuel rate
- gas rate
- inlet pressure
- oulet pressure
- H2S
- PSI

COMPRESSOR
- compressor name
- gas
- suction pressure
- flow

Since each object has unique properties associated with it, what would be the best table design for this scenario?

1) Create seperate table for each object
2) Create one table for objects and a child table for properties

Those are the two methods that i could come up with, any other suggestion would be appreciated.

[eoghain]

I can't answer what is better, since that all depends on what the data is going to be used for, but I can give you the guidelines that I use when designing a database.

1. Divide up my data into logical objects - looks like you've already done this

2. Find common data in all objects and see if it makes sence to store that data separately and link it to my objects.

3. Check all remaining fields in each object and see if they are all always needed (can any fields be null). If there are fields that can be null does it make sense to store them in the object or separately and link them in.
   [/list=1]
   Now looking at what you have, it seems to me that building a table for each of these separate objects would make sense. But without more information about the project and other data points that would be needed it's hard to say more.